Optimizing diagnostic methods and stem cell transplantation outcomes in pediatric bone marrow failure: a 50-year single center experience

Non-reversible bone marrow failure (BMF) in pediatric patients is caused by a broad spectrum of underlying diseases, including inherited bone marrow failure syndromes (IBMFS), (pre)malignant disease, and (idiopathic) aplastic anemia (AA) [1‐3]. In up to 50% of these patients, a germline genetic defect can be identified causing BMF [4‐6]. The etiology of IBMFS differs depending on the genetic defect or affected pathway. Well-characterized IBMFS include Fanconi Anemia (FA), caused by impaired DNA repair, Diamond-Blackfan Anemia (DBA) and Shwachman-Diamond Syndrome (SDS), associated with ribosome biogenesis defects, Severe Congenital Neutropenia (SCN) which is the result of deficient maturation of neutrophils and Dyskeratosis Congenita (DC) associated with inadequate telomere maintenance, resulting in shortened telomeres. In addition to cytopenia, patients with IBMFS often show multiorgan extra-hematological defects and have an increased risk for cancer, especially secondary myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) [1, 7]. Also independent of IBMFS, as a result of clonal hematopoiesis, MDS can develop and cause BMF [8]. Refractory cytopenia in childhood (RCC) is the most common subtype of MDS in children, characterized by less than 2% blasts in the peripheral blood and bone marrow [8‐10]. The remainder of patients with BMF with unknown etiology is most often diagnosed as idiopathic AA [4]. Although the exact mechanism of AA is unknown, it is generally accepted to be caused by immune dysregulation [11, 12].

Timely recognition of (irreversible) BMF and identification of the underlying cause of BMF result in reduced risks of invasive infections and bleeding complications. This requires guidelines for a consistent diagnostic approach for pediatric patients suspected of BMF. In these guidelines, in addition to morphological and histological examination of bone marrow aspirates and biopsies, molecular techniques and functional assays are of increasing value. In general, genetic screening is initiated in case of extra-hematological physical abnormalities and/or a positive family history [4, 13]. However, these aspects are not always identifiable at the onset of BMF, especially in young children. As a result, patients might be diagnosed as AA. Therefore, we proposed an extensive unbiased diagnostic algorithm [4]. This method includes broad genetic analysis and telomere length analysis in addition to the conventional diagnostic tests to achieve swift interpretation of first-line diagnostics and a higher yield of identifiable defects underlying irreversible BMF. As a result, the cause of BMF in pediatric patients was more often discovered. This allowed for risk-adapted organ and cancer monitoring, family counseling and prompt initiation of a curative treatment regime, mainly by allogeneic hematopoietic stem cell transplantation (HSCT, Fig. 1). In addition, the remaining group of BMF with unknown origin or etiology, classified as AA, became smaller. This provides the opportunity to stratify a group of AA for upfront immune suppressive therapy (IST) treatment with expected superior therapy efficacy and provides a better homogeneous group for future research to explore AA etiology and pathogenesis.

Pediatric BMF patients are transplanted in our center for over 50 years. The first severe AA (SAA) patient (BMF of unknown origin at the time) was transplanted in 1971 and the first patient with severe BMF due to FA was transplanted in 1972. During the years following, conditioning regimens developed from toxic myeloablative to more reduced intensity and toxicity regimes (Supplementary Table 1). Especially for the IBMFS group, chemotherapy regimens aimed at less organ toxicity. While patients transplanted by using bone marrow (BM) from HLA-identical sibling donors showed superior outcomes, delay in HSCT or second-line HSCT after upfront IST, both resulted in higher mortality in pediatric patients [14]. The introduction of post-transplantation cyclophosphamide (PT-Cy) to prevent graft versus host disease (GVHD) provided access to a great number of (mismatched) unrelated donors. In the following years, outcomes of HSCT by using (mismatched) unrelated donors, increasingly improved and are currently comparable with the success rates of HLA-matched transplantation.

Managing pediatric patients with BMF is challenging at all stages of the care pathway. An individualized treatment and surveillance plan will depend on the underlying cause of the BMF and the severity of the condition. Identification of the causative defects underlying BMF is particularly difficult in otherwise healthy children without extra-hematological signs and symptoms. The caveat there is that these patients are categorized as AA. Given the lack of specific disease markers for AA, the diagnosis remains a diagnosis per exclusion. Recent advances in genomic evaluation including whole exome next-generation sequencing and functional analysis such a telomere length analysis, provide the opportunity for an unbiased diagnostic approach [4]. Concordant genetic and functional analyses provide the opportunity to detect novel genetic variations and to define genetic variations of unknown significance (VUS). Identification of the cause of BMF delivers crucial information for treatment strategies and monitoring protocols on individual patient level. Increased diagnostic yield of identified causes of BMF also results in a smaller remnant group to be diagnosed as AA. The increased homogeneity of this group can result in better outcomes of IST and provides a better-defined group for future AA related research. In addition, the unbiased diagnostic approach provides the basic information for crucial insights in understanding DNA repair, telomere and ribosome biology, and hematopoietic stem cell and stromal niche regulation.

Risk assessment of malignant transformation represents another challenge at the diagnostic stage. Based on morphological and histological characteristic changes defined as dysplasia, efforts are directed towards the differentiation between hypoplastic bone marrow at risk for clonal evolution and malignant transformation (such as in MDS-RCC) and hypoplastic bone marrow without increased risk of malignant transformation (such as in SAA) [15‐17]. Also here, molecular diagnostics can be of crucial importance. Next to the evaluation of clonal hematopoiesis driven by well-known MDS/AML related cytogenetic alterations, somatic mutations might be predictive of malignant transformation [18, 19]. In addition, increasing evidence is available that clonal hematopoiesis does not equate malignant transformation on itself [20]. In particular situations, such as with monosomy 7 in SAMD-9 or GATA-2 germline mutation driven BMF, clonal hematopoiesis may even be part of an escape mechanism to restore hematopoiesis and should not require immediate definitive treatment to prevent malignant transformation. All in all, molecular bone marrow analysis beyond the frequent cytogenetic changes and expert opinion are required to identify markers of clonal evolution and malignant transformation for an individualized monitoring plan of BMF patients at risk.

Also, the treatment of pediatric BMF patients is accompanied with challenges. The majority of these patients present with or develop severe cytopenia or risk indicators for hematologic malignancies. Thereby for most patients, HSCT is an attractive and often the only curative treatment option. Based on our long-term experience in HSCT for pediatric patients with BMF as presented here, we draw the following conclusions. HSCT has an unequaled curative potential to restore BM defects independent of the underlying cause. Moreover, with the current successes of HLA-mismatched transplantations by effective GVHD prevention such as PT-Cy or alpha/beta T-cell depletion, and reduced toxicity conditioning regimens, HSCT is increasingly offered as the first choice of treatment to pediatric BMF patients. However, short- and long-term (event-free) survival rates are evidently dependent on the management of the patients before and after transplantation. A structured multidisciplinary approach directed towards unbiased but swift diagnostics at the suspicion to diagnosis stage together with maximal supportive care to prevent infections and bleeding result into significant gain of survival rates within the first 3 months after the onset of the disease. This is underlined by the significant improved overall and disease-free survival rates of BMF patients transplanted after 2017 independent of the underlying cause. In addition, especially in patients with IBMFS, protocolized approaches to systematically screen for late effects post HSCT are of crucial importance. The second decline in survival of these patients 10 to 20 years after HSCT, often explained by secondary malignancies, lung- and liver fibrosis and other organ failures, underlines the essence of personalized follow-up protocols. This wide variety of long-term toxicities including chronic GVHD, delayed or unbalanced immune reconstitution, iron overload, developmental delay and psychosocial impairment and malignant disease require a multidisciplinary approach guided by centers of expertise [21].